This analysis uses Google Trends information to analyze the queries made around the time of the August 21, 2017 solar eclipse.

We will look at correlations by state with Google searches for the phrases: "solar eclipse" & "eyes hurt"

Then we will look at an index of the 2 searchs by state

library(plotly)
library(gtrendsR)

SolEcl=gtrends("solar eclipse","US",time = "2017-08-01 2017-08-31")$interest_by_region
IsHurt<-gtrends("eyes hurt","US",time = "2017-08-20 2017-08-31")$interest_by_region

SolBld=merge(SolEcl, IsHurt, by=c("location","location"), all.x = TRUE)
names(SolBld)[names(SolBld) == 'hits.x'] <- 'SolEclInterest'
names(SolBld)[names(SolBld) == 'hits.y'] <- 'IsHurtInterest'
names(SolBld)[names(SolBld) == 'location'] <- 'state'

Unfortunately the path of the eclipse was not readily available. So, I had to do some manupulations in excel. The raw data is available from NASA here: https://eclipse.gsfc.nasa.gov/SEpath/SEpath2001/SE2017Aug21Tpath.html The central line was processed and saved as SolarEclipsePath.csv in the working directory.

Plotting out the Solar Eclipse interest you can see the states that queried this the most during the month of the eclipse.

map

Plotting out the Eyes Hurt interest you can see the states that queried this the most AFTER the eclipse.

map

Clearly the states along the central line had the highest interest in solor eclipse and painful eyes following the eclipse.

map